Site-specific Weed Management Research Articles

Weed species classification with UAV imagery and standard CNN models: Assessing the frontiers of training and inference phases

Accurate weed species identification is crucial for effective site-specific weed management (SSWM), enabling targeted and timely control measures for each weed in crop field. This study advanced the current approach to species-level weed identification during the early growth stage by integrating unmanned aerial vehicles (UAVs) imagery with standard convolutional neural networks (CNNs) models such as VGG16, Resnet152 and Inception-Resnet-v2. For this, a robust dataset was created with 33,467 labels of weeds (Atriplex patula, Chenopodium album, Convolvulus arvensis, Cyperus rotundus, Lolium rigidum, Portulaca oleracea, Salsola kali, Solanum nigrum) and crops (maize, tomato), which was subjected to different training, validation and test scenarios. Model inputs were adjusted in order to align them with the information represented by the UAV images. Initially, models were developed in balanced scenarios, gradually increasing label numbers to assess their performance. Inception-ResNet-v2 achieved over 90% accuracy with 400 labels, while ResNet152 and VGG16 required 600 and 800 labels, respectively, for similar accuracy. In a more complex and realistic scenarios with unbalanced datasets, Inception-ResNet-v2 outperformed, likely due to its deeper architecture and enhanced capability to capture intricate features and patterns within UAV images. The study emphasized the importance of the minority-to-majority species ratio in unbalanced datasets, which affects minority species classification. To prevent misclassification, it is crucial to determine the right number of labels for CNN model training and validation. Weed maps were generated after species classification using the Faster R–CNN algorithm as an object detector. This advancement in methodology facilitates the precise and efficient implementation of SSWM techniques.

Weed Management Using UAV and Remote Sensing in Malaysia Paddy Field: A Review

Controlling weed infestation is pivotal to achieving the maximum yield in paddy fields. At a time of exponential human population growth and depleting arable land mass, finding the solution to this problem is crucial. For a long time, herbicides have been the most favoured approach for weed control due to their efficacy and ease of application. However, adverse effects on the environment due to the excessive use of herbicides have prompted more cautious and effective herbicide usage. Many weed species tend to dominate the field, and the weed thrived in patches, rendering conventional broad herbicide spraying futile. Site-specific weed management (SSWM) consists of two strategies: weed mapping and selective herbicide application. Since its introduction into the agriculture sector, unmanned aerial vehicles (UAV) have become the platform of choice for carrying both the remote sensing system for weed mapping and the selective application of herbicide. Red-Green-Blue (RGB), multispectral and hyperspectral sensors on UAVs enable highly accurate weed mapping. In Malaysia, adopting this technology is highly possible, given the nature of government-administrated rice cultivation. This review provides insight into the weed management practice using remote sensing techniques on UAV platforms with potential applications in Malaysia's paddy field. It also discusses the recent works on weed mapping with imaging remote sensing on a UAV platform.

Spectral mixture analysis for weed traits identification under varying resolutions and growth stages

Spectral and spatial resolutions are critical factors in Site-Specific Weed Management (SSWM) applications, which aim to adjust and apply herbicides according to the weed composition and coverage during early growth. Thus, a reliable SSWM requires high spectral and spatial resolution data to create informative maps indicating and spatially localizing the weed species. Besides, upscaling and improving the mapping process are essential to motivate farmers to adopt SSWM practices. However, high-resolution data in spectral and spatial domains is not always available, and mixed pixels are highly probable. Therefore, evaluating approaches that deal with such pixels that pose a significant problem for weed classification tasks is vital. In this regard, Spectral Mixture Analysis (SMA) can exploit subpixel information from coarse-resolution data. However, despite its potential, SMA has not been examined for subpixel weed mapping. In this work, we aim to evaluate the potential advantages of SMA for coverage estimation of different weed traits (weed botanical group, photosynthetic pathway, and species). For this purpose, we first created a dataset containing weed species characterized by different botanical groups and photosynthetic pathways. Then, we evaluated the performance of SMA in estimating the weed’s coverage at varying growth stages and spectral and spatial resolutions. The experimental results revealed that using botanical weed groups as Endmembers (EM) is advantageous over other traits. A quantitative evaluation revealed strong correlation values between the actual and predicted weed coverage and a Mean Absolute Error (MAE) lower than 10 % for the different EM at all resolutions and growth stages. Besides, SMA outperformed the traditional classification-based approach in estimating weed coverage with up to ∼ 30 % lower Mean Absolute Error. Finally, our study establishes a basis for incorporating SMA in various SSWM practices.

Detection and mapping of Amaranthus spinosus L. in bermudagrass pastures using drone imagery and deep learning for a site‐specific weed management

AbstractWeed encroachment negatively affects pasture productivity by reducing herbage allowance, stocking rates, and livestock performance. Amaranthus spinosus L. is a weed species widely found in pastures worldwide and is considered challenging for ranchers due to its great potential for invasion, making it difficult to control. The high costs of chemical application and the global concern about environmental impacts restrict indiscriminate herbicide spraying in pastures. Site‐specific weed management (SSWM) is a weed management strategy based on weed spot‐spraying that has the potential to overcome these issues. Images from unmanned aerial vehicles (UAVs) can provide valuable information for weed mapping to drive the herbicide application in pastures. Deep learning techniques have been highlighted in image classification tasks. We developed a deep convolutional neural network (CNN)‐based image segmentation model based on the U‐Net architecture to detect and map Amaranthus spinosus in bermudagrass pastures using red–green–blue images acquired through UAV flying in moderate‐high altitude. The images were acquired from twelve paddocks under three treatments (weed‐free, weed‐strips, or weed‐infested) during the summer (2021‐2022). The CNN model was able to detect around 80% of the A. spinosus with an average prediction accuracy of 94%. Our weed mapping showed the potential of using the U‐Net model to generate a herbicide application map to be inserted into the sprayer system, reducing up to 76% of the amount of herbicide applied. Further studies are encouraged to increase the robustness of the model across species and development stages and develop sprayer systems to implement the spot‐spraying in field conditions.

Channel Attention GAN-Based Synthetic Weed Generation for Precise Weed Identification.

Weed is a major biological factor causing declines in crop yield. However, widespread herbicide application and indiscriminate weeding with soil disturbance are of great concern because of their environmental impacts. Site-specific weed management (SSWM) refers to a weed management strategy for digital agriculture that results in low energy loss. Deep learning is crucial for developing SSWM, as it distinguishes crops from weeds and identifies weed species. However, this technique requires substantial annotated data, which necessitates expertise in weed science and agronomy. In this study, we present a channel attention mechanism-driven generative adversarial network (CA-GAN) that can generate realistic synthetic weed data. The performance of the model was evaluated using two datasets: the public segmented Plant Seedling Dataset (sPSD), featuring nine common broadleaf weeds from arable land, and the Institute for Sustainable Agro-ecosystem Services (ISAS) dataset, which includes five common summer weeds in Japan. Consequently, the synthetic dataset generated by the proposed CA-GAN obtained an 82.63% recognition accuracy on the sPSD and 93.46% on the ISAS dataset. The Fréchet inception distance (FID) score test measures the similarity between a synthetic and real dataset, and it has been shown to correlate well with human judgments of the quality of synthetic samples. The synthetic dataset achieved a low FID score (20.95 on the sPSD and 24.31 on the ISAS dataset). Overall, the experimental results demonstrated that the proposed method outperformed previous state-of-the-art GAN models in terms of image quality, diversity, and discriminability, making it a promising approach for synthetic agricultural data generation.

Cross-domain transfer learning for weed segmentation and mapping in precision farming using ground and UAV images

Weed and crop segmentation is becoming an increasingly integral part of precision farming that leverages the current computer vision and deep learning technologies. Research has been extensively carried out based on images captured with a camera from various platforms. Unmanned aerial vehicles (UAVs) and ground-based vehicles including agricultural robots are the two popular platforms for data collection in fields. They all contribute to site-specific weed management (SSWM) to maintain crop yield. Currently, the data from these two platforms is processed separately, though sharing the same semantic objects (weed and crop). In our paper, we have proposed a novel method with a new deep learning-based model and the enhanced data augmentation pipeline to train field images alone and subsequently predict both field images and UAV images for weed segmentation and mapping. The network learning process is visualized by feature maps at shallow and deep layers. The results show that the mean intersection of union (IOU) values of the segmentation for the crop (maize), weeds, and soil background in the developed model for the field dataset are 0.744, 0.577, 0.979, respectively, and the performance of aerial images from an UAV with the same model, the IOU values of the segmentation for the crop (maize), weeds and soil background are 0.596, 0.407, and 0.875, respectively. To estimate the effect on the use of plant protection agents, we quantify the relationship between herbicide spraying saving rate and grid size (spraying resolution) based on the predicted weed map. The spraying saving rate is up to 90 % when the spraying resolution is at 1.78 × 1.78 cm2. The study shows that the developed deep convolutional neural network could be used to classify weeds from both field and aerial images and delivers satisfactory results. To achieve this performance, it is crucial to perform preprocessing techniques that reduce dataset differences between two distinct domains.

Agricultural weed identification in images and videos by integrating optimized deep learning architecture on an edge computing technology

Recent advancements in deep learning (DL)-based model optimization techniques have resulted in better weed identification accuracy. However, optimizing these models to identify weeds in images captured using small unmanned aerial system (UAS) has not been much explored. Moreover, leveraging the optimized model on resource constrained edge platform that could be easily integrated with UAS for real-time weed identification could be of significant advantage in developing precision aerial spraying weed management technology. Therefore, this study proposes YOLO-Spot model that is based on YOLOv7-tiny architecture, has been optimized and reconstructed to identify weeds amongst crop plants in aerial images and videos. The optimized model tends to use a smaller number of trainable parameters and reduced feature map sizes for weed identification. Most of the redundant convolutional layers along with feature map sizes have been reduced coupled with an integration of a novel module re-parameterized convolutional layer (RCL) within the neck component of the network. Furthermore, YOLO-Spot model has been trained on the three image resolutions, 320 × 320, 640 × 640 and 1280 × 1280, and has been named as YOLO-Spot_S, YOLO-Spot_M and YOLO-Spot_L, respectively. Out of all the variants, YOLO-Spot_M model has achieved significant prediction accuracy as compared to other variants and a denser layered model YOLOv7-Base. YOLO-Spot_M model utilizes over 75 % less parameters and 86 % reduced GFLOPs compared to YOLOv7-Base. As per the results, YOLO-Spot_M has outperformed YOLOv7-Base by achieving +1.3 % and +2.7 % overall accuracy and mAP(@0.5), respectively. The optimized architecture utilizes 4X less power (in W) when trained on a normal graphical processing unit (GPU). Moreover, converting YOLO-Spot_M to half-precision (FP16) for resource constrained device deployment (AGX Xavier), led to a +0.6 % accuracy and 5X faster weed recognition accuracy in aerial images and videos during inferencing. Based on the metrics obtained, YOLO-Spot_M model is recommended model that could be integrated with remote sensing technologies for site-specific weed management.

Transformer in UAV Image-Based Weed Mapping

Weeds affect crop yield and quality due to competition for resources. In order to reduce the risk of yield losses due to weeds, herbicides or non-chemical measures are applied. Weeds, especially creeping perennial species, are generally distributed in patches within arable fields. Hence, instead of applying control measures uniformly, precision weeding or site-specific weed management (SSWM) is highly recommended. Unmanned aerial vehicle (UAV) imaging is known for wide area coverage and flexible operation frequency, making it a potential solution to generate weed maps at a reasonable cost. Efficient weed mapping algorithms need to be developed together with UAV imagery to facilitate SSWM. Different machine learning (ML) approaches have been developed for image-based weed mapping, either classical ML models or the more up-to-date deep learning (DL) models taking full advantage of parallel computation on a GPU (graphics processing unit). Attention-based transformer DL models, which have seen a recent boom, are expected to overtake classical convolutional neural network (CNN) DL models. This inspired us to develop a transformer DL model for segmenting weeds, cereal crops, and ‘other’ in low-resolution RGB UAV imagery (about 33 mm ground sampling distance, g.s.d.) captured after the cereal crop had turned yellow. Images were acquired during three years in 15 fields with three cereal species (Triticum aestivum, Hordeum vulgare, and Avena sativa) and various weed flora dominated by creeping perennials (mainly Cirsium arvense and Elymus repens). The performance of our transformer model, 1Dtransformer, was evaluated through comparison with a classical DL model, 1DCNN, and two classical ML methods, i.e., random forest (RF) and k-nearest neighbor (KNN). The transformer model showed the best performance with an overall accuracy of 98.694% on pixels set aside for validation. It also agreed best and relatively well with ground reference data on total weed coverage, R2 = 0.598. In this study, we showed the outstanding performance and robustness of a 1Dtransformer model for weed mapping based on UAV imagery for the first time. The model can be used to obtain weed maps in cereals fields known to be infested by perennial weeds. These maps can be used as basis for the generation of prescription maps for SSWM, either pre-harvest, post-harvest, or in the next crop, by applying herbicides or non-chemical measures.

Analysis of Stable Diffusion-derived fake weeds performance for training Convolutional Neural Networks

Weeds challenge crops by competing for resources and spreading diseases, impacting crop yield and quality. Effective weed detection can enhance herbicide application, thus reducing environmental and health risks. A major challenge in Site-Specific Weed Management (SSWM) is developing a reliable weed identification system, especially given the diversity and similarity between certain weeds and crops during early growth stages. Image-based deep learning (DL) methods have become vital for weed classification. However, accurate weed classification and detection using DL techniques face the bottleneck of requiring large labeled data. Furthermore, labeling this specific extensive data is a time-consuming and tedious task apart from necessitating weed science experts. This research's central focus is to present a novel approach to weed detection using convolutional neural network (CNN) classifiers, specifically Yolov8l and RetinaNet, augmented with Stable Diffusion data i.e., artificial weed images. Stable Diffusion enhanced the training data, increasing the classifiers' adaptability. The study targeted specific weeds (Solanum nigrum L.; Portulaca oleracea L.; Setaria Verticillata L.) found in tomato crops, using a limited number of real images (30 samples) to produce artificial training images for the CNNs. All validation and test sets are comprised of real weed images. Results showed high performance when using only artificial images in terms of Mean Average Precision (mAP). In isolated conditions (0.91 mAP), i.e., only one weed species per image, an average performance gain of about 3% in all tests is obtained. When adding the artificial images to the real ones (mixed dataset), a mAP of 0.99 is obtained. In contrast, results using only artificial images obtained 0.81 mAP when detecting more than a single weed species. However, when implementing the trained CNNs with a mixed dataset, a 6% − 9% performance gain was achieved in all cases. A mAP of up to 0.93 was achieved in the most challenging conditions where weed species could overlap. The results indicate that the proposed approach outperformed existing methods, such as Generative Adversarial Networks (GANs) regarding mAP. Furthermore, the Yolov8l model distinctly emerged as the most favorable option for real-time detection systems considering Frame Detection Speed (FDS). Specifically, the Yolov8l model registered an FDS of 10.2 ms, which is considerably faster when compared to the 21.2 ms that the RetinaNet model exhibited. Additionally, the method is versatile and applicable to various crops and weed species, thereby enhancing automated weed management systems. This research illustrates that Stable Diffusion can efficiently expand small image sets, significantly reducing field imaging. The study offers valuable insights for future SSWM efforts utilizing artificially generated images for weed detection and classification.

Improved weed segmentation in UAV imagery of sorghum fields with a combined deblurring segmentation model

BackgroundEfficient and site-specific weed management is a critical step in many agricultural tasks. Image captures from drones and modern machine learning based computer vision methods can be used to assess weed infestation in agricultural fields more efficiently. However, the image quality of the captures can be affected by several factors, including motion blur. Image captures can be blurred because the drone moves during the image capturing process, e.g. due to wind pressure or camera settings. These influences complicate the annotation of training and test samples and can also lead to reduced predictive power in segmentation and classification tasks.ResultsIn this study, we propose DeBlurWeedSeg, a combined deblurring and segmentation model for weed and crop segmentation in motion blurred images. For this purpose, we first collected a new dataset of matching sharp and naturally blurred image pairs of real sorghum and weed plants from drone images of the same agricultural field. The data was used to train and evaluate the performance of DeBlurWeedSeg on both sharp and blurred images of a hold-out test-set. We show that DeBlurWeedSeg outperforms a standard segmentation model that does not include an integrated deblurring step, with a relative improvement of 13.4 % in terms of the Sørensen-Dice coefficient.ConclusionOur combined deblurring and segmentation model DeBlurWeedSeg is able to accurately segment weeds from sorghum and background, in both sharp as well as motion blurred drone captures. This has high practical implications, as lower error rates in weed and crop segmentation could lead to better weed control, e.g. when using robots for mechanical weed removal.

Leveraging machine learning and drone multispectral data for site-specific weed management in tomato agricultural areas

Leveraging machine learning and drone multispectral data for site-specific weed management in tomato agricultural areas

Trouble in the rice field: Distribution ecology and indicator value of weed species in the rice fields of Himalayan region

Field-based evaluation of indicator weed species in agriculture systems is effective for developing site-specific weed management. Because of anthropogenic activities and climate change, new weeds are invading agricultural fields. The current study examined the weed species in agricultural fields with a focus on diversity, distribution ecological patterns, and identification of indicator species, all of which might help in effective weed management practices. Following the calculation of the Importance Value Index for each weed species, ordination, and indicator species analysis were conducted. We collected vegetation data through random sampling, and the current study recorded 47 species, taxonomically distributed across 41 genera and 19 families. Asteraceae with 13% of species was the leading family followed by Poaceae, and Brassicaceae (11% each). Annual weed species with 49% were the dominant life span category. The phytogeographical analysis revealed that the highest number (28 species) recorded from the study area were native, while 19 species were alien. Of the total alien species, 4 were invasive, and 14 were naturalised and one species Bassia scoparia was casual. The diversity indices show significant differences (p ≤ 0.05) in weeds in rice fields recorded from the study area. Detrended correspondence analysis grouped the species Geranium rotundifolium, Sisymbrium loeselii, Mentha longifolia, Rumex dentatus, Amaranthus viridis on the extreme left side suggest a low score on axis 1 and a high score on axis 2. The analysis of indicator species revealed a significant difference (p ≤ 0.05) in agriculture fields across the sampled sites. According to the indicator species analysis, Poa annua, Persicaria hydropiper, Ranunculus sceleratus, and Echinochloa colona were the key weed species found in the study area. This investigation might aid weed management strategies to minimize crop losses and increase of rice production to fulfill the current food requirement.

Black-grass (Alopecurus myosuroides) in cereal multispectral detection by UAV

AbstractSite-specific weed management (on the scale of a few meters or less) has the potential to greatly reduce pesticide use and its associated environmental and economic costs. A prerequisite for site-specific weed management is the availability of accurate maps of the weed population that can be generated quickly and cheaply. Improvements and cost reductions in unmanned aerial vehicles (UAVs) and camera technology mean these tools are now readily available for agricultural use. We used UAVs to collect aerial images captured in both RGB and multispectral formats of 12 cereal fields (wheat [Triticum aestivumL.] and barley [Hordeum vulgareL.]) across eastern England. These data were used to train machine learning models to generate prediction maps of locations of black-grass (Alopecurus myosuroidesHuds.), a prolific weed in UK cereal fields. We tested machine learning and data set resampling methods to obtain the most accurate system for predicting the presence and absence of weeds in new out-of-sample fields. The accuracy of the system in predicting the absence ofA. myosuroidesis 69% and its presence above 5 g in weight with 77% accuracy in new out-of-sample fields. This system generates prediction maps that can be used by either agricultural machinery or autonomous robotic platforms for precision weed management. Improvements to the accuracy can be made by increasing the number of fields and samples in the data set and the length of time over which data are collected to gather data across the entire growing season.

Performance evaluation of YOLO v5 model for automatic crop and weed classification on UAV images

For sustainability and efficiency in maintaining high crop yield and less chemically polluted agricultural lands, precise weed mapping is essential for the total implementation of site-specific weed management which currently stands as a major challenge in present day agriculture. In this research, the robustness of the training epochs of You Only Look Once (YOLO) v5s, a Convolutional Neural Network (CNN) model was evaluated for the development of an automatic crop and weeds classification using UAV images. The images were annotated using a bounding box and they were trained on google colaboratory over 100, 300, 500, 600, 700 and 1000 epochs. The model detected and categorized five different classes which are sugarcane (Saccharum officinarum), banana trees (Musa), spinach (Spinacia oleracea), pepper (Capsicum), and weeds. To find the optimal performance on the test set, the model was trained across several epochs, and training was stopped when the test performance (classification accuracy, precision, and recall) began to drop. The obtained result shows that the performance of the classifier improved significantly as the range of training epochs tends to rise from 100 through to 600 epochs. Meanwhile, a slight decline was observed as the number of epoch was increased to 700 when the classification accuracy, the precision of weed and recall of 65, 43 and 43%, respectively, was recorded as against 67, 78 and 34% that was obtained as the classification accuracy, weed precision and recall, respectively, at 600 epochs. This decline continued even when the epoch was increased to 1000 where classification accuracy, weed precision and recall of 65%, 45% and 40%, respectively was obtained. The results showed that the training epoch of YOLOv5s significantly affects the model's robustness in automatic crop and weep classification and identified 600 as the epoch for optimal performance.

Improving agricultural robot patch-spraying accuracy and precision through combined error adjustment

Site-specific weed management (SSWM) involves precision spraying operations based on the spatial variability of weeds in the field. High spraying precision enables extensive economic and environmental benefits. We developed a high-precision patch-spraying robot system based on weed maps to conduct patch-spraying operations at the individual plant scale. The patch-spraying droplet deposition effect is evaluated by post-spraying droplet deposition images. The agricultural robot mainly consists of a walking chassis, spraying, and control systems. The walking chassis provides walking power and support for the suspended spraying system. Combining the static structural and dynamic error models of the robot improves patch-spraying accuracy and precision. Twelve patch-spraying simulations were carried out on the leveling site. Different correction methods and driving speeds were tested, with driving rates of 0.5, 1.0, and 1.5 m/s. Aerial images of the patch-spraying process were evaluated using pixel point-based discriminant statistics after operation completion. The droplet deposition coverage accuracy ranged from 95.9 to 98.8 %. The center of droplet deposition offset distance was 3.7–7.2 cm. Field results showed that the patch-spraying weed control rate based on individual plant scales (spraying precision ≤ 15 cm) was 94.2–98.1 %. In summary, the developed agricultural robot performs patch-spraying operations at the individual plant scale. Patch-spraying accuracy and precision are improved by modeling corrections for random errors. A general error correction method for the deposition position of fog droplets is explored to achieve precise droplet delivery in the walking state of plant protection operation equipment.

Multispectral UAV data for detection of weeds in a citrus farm using machine learning and Google Earth Engine: Case study of Morocco

Accurate and timely weed mapping between and within trees is considered one of the major challenges in the site-specific weed management systems. This research presents the first cloud computing approach based on the incorporation of multispectral Unmanned Aerial Vehicle (UAV) imagery in the Google Earth Engine (GEE) programming environment with the aim of improving the mapping of weed patches between and within trees in a citrus farm. For this purpose, the UAV multispectral bands (red, green, blue, near infrared, and red edge), as well as the estimated vegetation height from the UAV Digital Elevation Model (DEM) were analysed in terms of tree and weed discrimination and used as input into Random Forest (RF) and k-nearest neighbors (KNN) machine learning algorithms. From the DEM, the Digital Terrain Model (DTM) was estimated using the Inverse Distance Weighted Interpolation (IDW) of the elevation values of dense points on the soil. The plant height was derived by subtracting the DTM from the DEM, resulting in the Canopy Height Model (CHM). The experimental results show that: (i) the combination of spectral bands and the CHM can classify both trees and weeds with an overall accuracy reached 96.87%; (ii) the RF classifier was more robust compared to KNN in the classification performance; (iii) when compared to the use of UAV spectral bands, the addition of the CHM can improve the accuracy of crop classification by 13.36% (KNN) and 1.79% (RF). Furthermore, the integration of UAV imagery in the GEE was highly efficient in terms of automation of the UAV imagery processing.

Early-Season Mapping of Johnsongrass (Sorghum halepense), Common Cocklebur (Xanthium strumarium) and Velvetleaf (Abutilon theophrasti) in Corn Fields Using Airborne Hyperspectral Imagery

Accurate information on the spatial distribution of weeds is the key to effective site-specific weed management and the efficient and sustainable use of weed control measures. This work focuses on the early detection of johnsongrass, common cocklebur and velvetleaf present in a corn field using high resolution airborne hyperspectral imagery acquired when corn plants were in a four to six leaf growth stage. Following the appropriate radiometric and geometric corrections, two supervised classification techniques, such as spectral angle mapper (SAM) and spectral mixture analysis (SMA) were applied. Two different procedures were compared for endmember selections: field spectral measurements and automatic methods to identify pure pixels in the image. Maps for both, overall weeds and for each of the three weed species, were obtained with the different classification methods and endmember sources. The best results were achieved by defining the endmembers through spectral information collected with a field spectroradiometer. Overall accuracies ranged between 60% and 80% using SAM for maps that do not differentiate the weed species while it decreased to 52% when the three weed species were individually classified. In this case, the SMA classification technique clearly improved the SAM results. The proposed methodology shows it to be a promising prospect to be applicable to low cost images acquired by the new generation of hyperspectral sensors onboard unmanned aerial vehicles (UAVs).

Deep Object Detection of Crop Weeds: Performance of YOLOv7 on a Real Case Dataset from UAV Images

Weeds are a crucial threat to agriculture, and in order to preserve crop productivity, spreading agrochemicals is a common practice with a potential negative impact on the environment. Methods that can support intelligent application are needed. Therefore, identification and mapping is a critical step in performing site-specific weed management. Unmanned aerial vehicle (UAV) data streams are considered the best for weed detection due to the high resolution and flexibility of data acquisition and the spatial explicit dimensions of imagery. However, with the existence of unstructured crop conditions and the high biological variation of weeds, it remains a difficult challenge to generate accurate weed recognition and detection models. Two critical barriers to tackling this challenge are related to (1) a lack of case-specific, large, and comprehensive weed UAV image datasets for the crop of interest, (2) defining the most appropriate computer vision (CV) weed detection models to assess the operationality of detection approaches in real case conditions. Deep Learning (DL) algorithms, appropriately trained to deal with the real case complexity of UAV data in agriculture, can provide valid alternative solutions with respect to standard CV approaches for an accurate weed recognition model. In this framework, this paper first introduces a new weed and crop dataset named Chicory Plant (CP) and then tests state-of-the-art DL algorithms for object detection. A total of 12,113 bounding box annotations were generated to identify weed targets (Mercurialis annua) from more than 3000 RGB images of chicory plantations, collected using a UAV system at various stages of crop and weed growth. Deep weed object detection was conducted by testing the most recent You Only Look Once version 7 (YOLOv7) on both the CP and publicly available datasets (Lincoln beet (LB)), for which a previous version of YOLO was used to map weeds and crops. The YOLOv7 results obtained for the CP dataset were encouraging, outperforming the other YOLO variants by producing value metrics of 56.6%, 62.1%, and 61.3% for the mAP@0.5 scores, recall, and precision, respectively. Furthermore, the YOLOv7 model applied to the LB dataset surpassed the existing published results by increasing the mAP@0.5 scores from 51% to 61%, 67.5% to 74.1%, and 34.6% to 48% for the total mAP, mAP for weeds, and mAP for sugar beets, respectively. This study illustrates the potential of the YOLOv7 model for weed detection but remarks on the fundamental needs of large-scale, annotated weed datasets to develop and evaluate models in real-case field circumstances.

Aerial-Based Weed Detection Using Low-Cost and Lightweight Deep Learning Models on an Edge Platform

Highlights Lightweight deep learning models were trained on an edge device to identify weeds in aerial images. A customized configuration file was setup to train the models. These models were deployed to detect weeds in aerial images and videos (near real-time). CSPMobileNet-v2 and YOLOv4-lite are recommended models for weed detection using edge platform. Abstract. Deep learning (DL) techniques have proven to be a successful approach in detecting weeds for site-specific weed management (SSWM). In the past, most of the research work has trained and deployed pre-trained DL models on high-end systems coupled with expensive graphical processing units (GPUs). However, only a limited number of research studies have used DL models on an edge system for aerial-based weed detection. Therefore, while focusing on hardware cost minimization, eight DL models were trained and deployed on an edge device to detect weeds in aerial-image context and videos in this study. Four large models, namely CSPDarkNet-53, DarkNet-53, DenseNet-201, and ResNet-50, along with four lightweight models, CSPMobileNet-v2, YOLOv4-lite, EfficientNet-B0, and DarkNet-Ref, were considered for training a customized DL architecture. Along with trained model performance scores (average precision score, mean average precision (mAP), intersection over union, precision, and recall), other model metrics to assess edge system performance such as billion floating-point operations/s (BFLOPS), frame rates/s (FPS), and GPU memory usage were also estimated. The lightweight CSPMobileNet-v2 and YOLOv4-lite models outperformed others in detecting weeds in aerial image context. These models were able to achieve a mAP score of 83.2% and 82.2%, delivering an FPS of 60.9 and 61.1 during near real-time weed detection in aerial videos, respectively. The popular ResNet-50 model achieved a mAP of 79.6%, which was the highest amongst all the large models deployed for weed detection tasks. Based on the results, the two lightweight models, namely, CSPMobileNet-v2 and YOLOv4-lite, are recommended, and they can be used on a low-cost edge system to detect weeds in aerial image context with significant accuracy. Keywords: Aerial image, Deep learning, Edge device, Precision agriculture, Weed detection.

TIA-YOLOv5: An improved YOLOv5 network for real-time detection of crop and weed in the field.

Development of weed and crop detection algorithms provides theoretical support for weed control and becomes an effective tool for the site-specific weed management. For weed and crop object detection tasks in the field, there is often a large difference between the number of weed and crop, resulting in an unbalanced distribution of samples and further posing difficulties for the detection task. In addition, most developed models tend to miss the small weed objects, leading to unsatisfied detection results. To overcome these issues, we proposed a pixel-level synthesization data augmentation method and a TIA-YOLOv5 network for weed and crop detection in the complex field environment. The pixel-level synthesization data augmentation method generated synthetic images by pasting weed pixels into original images. In the TIA-YOLOv5, a transformer encoder block was added to the backbone to improve the sensitivity of the model to weeds, a channel feature fusion with involution (CFFI) strategy was proposed for channel feature fusion while reducing information loss, and adaptive spatial feature fusion (ASFF) was introduced for feature fusion of different scales in the prediction head. Test results with a publicly available sugarbeet dataset showed that the proposed TIA-YOLOv5 network yielded an F1-scoreweed, APweed and mAP@0.5 of 70.0%, 80.8% and 90.0%, respectively, which was 11.8%, 11.3% and 5.9% higher than the baseline YOLOv5 model. And the detection speed reached 20.8 FPS. In this paper, a fast and accurate workflow including a pixel-level synthesization data augmentation method and a TIA-YOLOv5 network was proposed for real-time weed and crop detection in the field. The proposed method improved the detection accuracy and speed, providing very promising detection results.